Pre-transplant and persistent post-transplant hyperparathyroidism and kidney transplant outcomes: a systematic review and meta-analysis Authors: Daniele Vetrano, Simona Barbuto, Francesco Aguanno, Paolo Mastromauro, Valeria Grandinetti, Giorgia Comai, Gaetano La Manna, Giuseppe Cianciolo Published: 2026 ABSTRACT ABSTRACTBackgroundHyperparathyroidism (HPT) is common in chronic kidney disease and may persist after kidney transplantation, but its prognostic significance remains uncertain because of heterogeneous definitions and variable timing of assessment.MethodsWe performed a systematic review and meta-analysis of observational studies in adult kidney transplant recipients (KTRs) in accordance with PRISMA 2020 (PROSPERO CRD420261293712). PubMed/MEDLINE and Embase were searched from inception to 9 February 2026. Exposures included pre- and post-transplant parathyroid hormone (PTH)/Persistent HPT assessed as categorical or continuous variables. Outcomes were all-cause mortality, overall graft failure, and death-censored graft failure. Random-effects models were applied, with Hartung–Knapp sensitivity analyses; risk of bias was assessed using QUIPS.Results23 studies were included in the systematic review, and 19 studies (26 266 KTRs) contributed to the meta-analyses (3 pre-transplant studies N = 12 819; 16 post-transplant studies N = 13 447). Pre-transplant HPT was not associated with all-cause mortality (HR 0.80) and showed an association with death-censored graft failure (DCGF) in the main analysis (HR 1.42) that lost statistical significance with Hartung–Knapp adjustment. By contrast, categorical Persistent HPT was associated with higher risks of all-cause mortality (HR 1.74), overall graft failure (HR 2.16), and DCGF (HR 1.92), with consistent Hartung–Knapp results. Phenotype-stratified analyses showed a gradient of graft failure risk from normocalcemic HPT (HR 1.66) to hypercalcemic HPT (HR 2.67), with minimal heterogeneity. Continuous PTH analyses showed substantial heterogeneity and were not robust in sensitivity analyses.ConclusionsPost-transplant Persistent HPT is a consistent risk marker for mortality and adverse graft outcomes, whereas the prognostic significance of pre-transplant HPT remains to be clarified. Standardized definitions and prospective studies are needed to clarify causality and treatment effects.Keywords: CKD-MBD, hypercalcemia, kidney transplantation, meta-analysis, persistent hyperparathyroidism What was known: Persistent post-transplant hyperparathyroidism (HPT) is frequent in kidney transplant recipients (KTRs) and reflects ongoing Chronic Kidney Disease-Mineral Bone Disorders (CKD-MBD) activity, with plausible links to cardiovascular burden and allograft injury; however, observational studies have reported inconsistent associations with mortality and graft outcomes, largely due to heterogeneous exposure definitions, timing of assessment, and endpoint selection. Persistent post-transplant hyperparathyroidism (HPT) is frequent in kidney transplant recipients (KTRs) and reflects ongoing Chronic Kidney Disease-Mineral Bone Disorders (CKD-MBD) activity, with plausible links to cardiovascular burden and allograft injury; however, observational studies have reported inconsistent associations with mortality and graft outcomes, largely due to heterogeneous exposure definitions, timing of assessment, and endpoint selection. Secondary HPT may promote vascular and soft-tissue calcification and reflect long-standing CKD-MBD severity; it can also persist after transplantation, evolving into persistent post-transplant hyperparathyroidism (PHPT) and potentially amplifying downstream cardiovascular and allograft injury pathways. However, evidence on its independent impact on post-transplant outcomes has remained fragmented across cohorts and definitions. Secondary HPT may promote vascular and soft-tissue calcification and reflect long-standing CKD-MBD severity; it can also persist after transplantation, evolving into persistent post-transplant hyperparathyroidism (PHPT) and potentially amplifying downstream cardiovascular and allograft injury pathways. However, evidence on its independent impact on post-transplant outcomes has remained fragmented across cohorts and definitions. Given the high prevalence of Persistent HPT, the plausible continuum from pre- to post-transplant HPT, and conflicting observational findings, a systematic review and meta-analysis was needed to integrate the evidence, harmonize results across study-specific definitions, and quantify associations with mortality, graft failure, and death-censored graft failure, including phenotype-stratified estimates. Given the high prevalence of Persistent HPT, the plausible continuum from pre- to post-transplant HPT, and conflicting observational findings, a systematic review and meta-analysis was needed to integrate the evidence, harmonize results across study-specific definitions, and quantify associations with mortality, graft failure, and death-censored graft failure, including phenotype-stratified estimates. This study adds: Categorical Persistent HPT was consistently associated with higher risks of all-cause mortality, graft failure, and death-censored graft failure in pooled analyses, and the direction and magnitude of effects remained stable across sensitivity analyses, supporting a robust adverse prognostic signal beyond single-study findings. Categorical Persistent HPT was consistently associated with higher risks of all-cause mortality, graft failure, and death-censored graft failure in pooled analyses, and the direction and magnitude of effects remained stable across sensitivity analyses, supporting a robust adverse prognostic signal beyond single-study findings. Pre-transplant HPT was mainly associated with graft outcomes, while pooled estimates did not support a clear association with all-cause mortality. Pre-transplant HPT was mainly associated with graft outcomes, while pooled estimates did not support a clear association with all-cause mortality. Phenotype stratification identified a graded association with graft failure: normocalcemic HPT carried a moderate excess risk, whereas hypercalcemic forms showed larger effect estimates; heterogeneity was minimal in phenotype-based models, indicating greater consistency and prognostic value when calcium status is incorporated. Phenotype stratification identified a graded association with graft failure: normocalcemic HPT carried a moderate excess risk, whereas hypercalcemic forms showed larger effect estimates; heterogeneity was minimal in phenotype-based models, indicating greater consistency and prognostic value when calcium status is incorporated. Potential impact: Supports systematic monitoring of parathyroid hormone and calcium status both pre- and post-transplant, using phenotype-informed stratification to identify KTRs at higher risk for adverse graft and patient outcomes and to prioritize early multidisciplinary management across the transplant pathway. Supports systematic monitoring of parathyroid hormone and calcium status both pre- and post-transplant, using phenotype-informed stratification to identify KTRs at higher risk for adverse graft and patient outcomes and to prioritize early multidisciplinary management across the transplant pathway. Provides quantitative evidence to justify trials testing whether active correction of Persistent HPT, particularly hypercalcemic forms, reduces mortality and improves graft outcomes. These results inform patient selection (phenotype-based enrichment), choice of endpoints and timing of intervention in future studies. Provides quantitative evidence to justify trials testing whether active correction of Persistent HPT, particularly hypercalcemic forms, reduces mortality and improves graft outcomes. These results inform patient selection (phenotype-based enrichment), choice of endpoints and timing of intervention in future studies. Reinforces the need for standardized, time-anchored operational definitions of post-transplant HPT (thresholds, persistence, and calcium-based phenotypes) across clinical pathways, registries, and guidelines to improve comparability between studies, enable consistent risk stratification, and support evidence-based management decisions. Reinforces the need for standardized, time-anchored operational definitions of post-transplant HPT (thresholds, persistence, and calcium-based phenotypes) across clinical pathways, registries, and guidelines to improve comparability between studies, enable consistent risk stratification, and support evidence-based management decisions. INTRODUCTION Hyperparathyroidism (HPT) is a common complication in patients with chronic kidney disease (CKD) and, along with alterations in calcium, phosphate, and vitamin D, is a hallmark of Chronic Kidney Disease-Mineral Bone Disorders (CKD-MBD) [1]. These disturbances may lead to both skeletal and cardiovascular complications of CKD, including increased fracture risk (CKD-associated osteoporosis) and cardiovascular disease [2]. As kidney function declines, CKD-MBD abnormalities, including secondary HPT, become progressively more prevalent across disease stages, affecting roughly 40%–50% of patients by stage 3 and ultimately reaching up to 80%–90% of patients with stage 5, including those on dialysis [3, 4]. Kidney transplantation is the preferred kidney replacement therapy for eligible patients [5]. Although restoration of kidney function improves phosphate handling and vitamin D metabolism, abnormalities in mineral metabolism may not fully resolve and can persist after transplantation [6]. Parathyroid hormone (PTH) levels typically decline markedly within the first 3 months post-transplant; however, in a subset of patients, PTH remains elevated with or without hypercalcemia [7]. Reported rates of transient or persistent post-transplant HPT vary widely, ranging from 10% to 70% [8–10]. This variability is largely driven by differences in diagnostic definitions (including PTH thresholds) and the timing of post-transplant assessments. In addition, distinguishing Persistent HPT from de novo secondary HPT is often challenging, as the latter may reflect incomplete recovery or subsequent deterioration of graft function. Several studies have suggested that Persistent HPT may be associated with an increased risk of graft loss, mortality, and, in some reports, fractures [9–12]. Some authors suggested that Persistent HPT may induce nephrocalcinosis in allografts, with calcium crystal deposition and increased inflammation [13–15]. Moreover, PTH has been hypothesized to contribute to graft injury through hemodynamic mechanisms, including effects on renal arteriolar tone, renin secretion, and vascular stiffness. [16]. Finally, PTH-driven disturbances in calcium–phosphate balance and bone turnover may promote vascular calcification and left ventricular remodeling [17]. Multiple transplant cohorts have evaluated the association between pre- and post-transplant HPT and hard outcomes such as all-cause mortality, overall graft failure, and death-censored graft failure (DCGF), with signals generally suggesting a higher risk among patients with HPT [9–11, 18–37]. The literature remains challenging to synthesize because studies differ greatly in (i) the definition of HPT, (ii) the timing of PTH assessment, and (iii) the reported endpoint (such as overall graft failure, DCGF, or composite outcomes). Additionally, studies vary in the scale used for continuous PTH measurements (e.g. log-transformed, per SD, or per unit), which hampers direct comparison. Given this heterogeneity and the resulting uncertainty about risk stratification and phenotype-guided management, we conducted a systematic review and meta-analysis of pre- and post-transplant PTH/HPT and clinical outcomes. We analyzed categorical and continuous definitions separately, evaluated all-cause mortality, overall graft failure, and DCGF as distinct endpoints, and conducted phenotype-based sensitivity analyses (normocalcemic vs hypercalcemic Persistent HPT). By using clinically interpretable comparisons and explicitly addressing heterogeneity, we aimed to clarify the strength and consistency of these associations and to highlight where evidence remains uncertain. MATERIALS AND METHODS Systematic review This systematic review and meta-analysis were conducted according to PRISMA 2020 [33]. The protocol was registered in PROSPERO (ID: CRD420261293712). We searched PubMed/MEDLINE and Embase from inception to 9 February 2026, using terms related to kidney transplantation, parathyroid hormone/HPT, and clinical outcomes. Records were de-duplicated and screened independently by two reviewers (DV and SB at title/abstract and full-text stages, with disagreements resolved by consensus or a third reviewer [GC]). We included observational studies of adult kidney transplant recipients (KTRs) evaluating pre- and/or post-transplant PTH/HPT—modeled categorically (threshold-based) and/or continuously—and reporting at least one of the following outcomes: all-cause mortality, overall graft failure, or DCGF, with extractable hazard ratios (HRs) (or equivalent) and 95% confidence intervals (CI). Data were extracted in duplicate using a standardized form; when multiple models were available, the most fully adjusted estimate was preferentially used. Risk of bias was assessed independently by two reviewers using quality in prognosis studies (QUIPS) (Supplementary Fig. S1). The study selection process is shown in Fig. 1 (PRISMA flow diagram). PRISMA flow diagram. Exposure and outcome definitions Exposure was classified according to how post-transplant parathyroid status was operationalized in each study. Because definitions of HPT—particularly in the post-transplant setting—are not uniform across the literature, we adhered to the study-specific classifications and thresholds adopted by each author. Categorical HPT (overall): any study-defined threshold or category indicating pre- or post-transplant HPT (PRE: high group vs low group; POST: Persistent HPT vs reference group, such as no HPT, normalized PTH, or a lower category, as defined by each study). Categorical HPT (overall): any study-defined threshold or category indicating pre- or post-transplant HPT (PRE: high group vs low group; POST: Persistent HPT vs reference group, such as no HPT, normalized PTH, or a lower category, as defined by each study). Phenotype-specific categorical definitions (sensitivity):Normocalcemic Persistent HPT: persistent elevated PTH with normocalcemia.Hypercalcemic Persistent HPT: persistent elevated PTH with hypercalcemia or study-defined tertiary HPT. Phenotype-specific categorical definitions (sensitivity): Normocalcemic Persistent HPT: persistent elevated PTH with normocalcemia. Normocalcemic Persistent HPT: persistent elevated PTH with normocalcemia. Hypercalcemic Persistent HPT: persistent elevated PTH with hypercalcemia or study-defined tertiary HPT. Hypercalcemic Persistent HPT: persistent elevated PTH with hypercalcemia or study-defined tertiary HPT. Continuous PTH (exploratory): PTH modeled as a continuous variable, analyzed on the scale reported by each study. Continuous PTH (exploratory): PTH modeled as a continuous variable, analyzed on the scale reported by each study. We evaluated time-to-event outcomes reported in KTRs, extracting hazard ratios (HRs) from Cox (or equivalent) models. Outcomes were harmonized as follows: All-cause mortality: death from any cause during follow-up. All-cause mortality: death from any cause during follow-up. Overall graft failure: composite outcome including graft failure or death with a functioning graft. Graft failure was defined according to each study, typically as return to dialysis, re-transplantation, and/or graft nephrectomy. When a study reported only DCGF, that estimate was considered eligible only if no overall graft failure estimate was available, to avoid double-counting. Overall graft failure: composite outcome including graft failure or death with a functioning graft. Graft failure was defined according to each study, typically as return to dialysis, re-transplantation, and/or graft nephrectomy. When a study reported only DCGF, that estimate was considered eligible only if no overall graft failure estimate was available, to avoid double-counting. DCGF: graft failure excluding death with a functioning graft, i.e. censoring at death, as defined by each study. DCGF: graft failure excluding death with a functioning graft, i.e. censoring at death, as defined by each study. Exposure–outcome pairings were prespecified according to the available literature and analyzed separately. Specifically, pre-transplant categorical HPT was evaluated against all-cause mortality and DCGF; post-transplant overall categorical Persistent HPT against all-cause mortality, overall graft failure, and DCGF; phenotype-specific post-transplant Persistent HPT (normocalcemic and hypercalcemic) against overall GF; and continuous post-transplant PTH against overall graft failure. Statistical methods For each eligible study and outcome, we extracted the most fully adjusted HR and its 95% CI when available; otherwise, unadjusted estimates were used. HRs were meta-analyzed on the log scale. Standard errors (SE) were derived from 95% CIs as: SE = [ln(CI_upper) − ln(CI_lower)]/(2 × 1.96). To prevent unit-of-analysis errors, one effect estimate per study per meta-analysis was included. For the graft failure analysis, overall graft failure was preferred; if unavailable, DCGF was used as a proxy according to the prespecified rule above. Meta-analyses were conducted separately by exposure definition, outcome, and HPT phenotype (overall Persistent HPT; normocalcemic and hypercalcemic HPT as sensitivity analyses). Pooled estimates were obtained using the inverse-variance random-effects model, with between-study variance (τ²) estimated by restricted maximum likelihood (REML). As sensitivity analyses, random-effects CI were recalculated using the Hartung–Knapp method. Between-study heterogeneity was assessed using Cochran’s Q (with corresponding P-value), I², and τ² (with 95% CI for τ² based on the Q-profile method as implemented in the software). Given the limited number of studies in several comparisons, formal assessment of small-study effects/publication bias was not emphasized. All tests were two-sided, with P < 0.05 considered statistically significant. Analyses were performed in R (version 4.5.2; R Foundation for Statistical Computing, Vienna, Austria) using the meta package. RESULTS A total of 23 studies were selected through the systematic review process; of these, 19 studies including 26 266 KTRs were included in the meta-analysis [9–11, 18–37]. Overall, three pre-transplant studies contributed 12 819 participants, while 16 post-transplant studies contributed 13 447 participants. The overall characteristics of the included studies are summarized in Table 1. Table 1:Characteristics of the studies included in the systematic review.Pre/post KTxPost-KTx timingPTH assayHPT threshold N HPT prevalenceBaseline eGFR (ml/min/1.73 m2)OutcomeCovariate adjustmentIncluded in final analysisAraujo et al. (2018) [18]Post12 mNR100 pg/ml91162%56 (44–67)Death-censored graft failurePHPT: HR 1.99, 95% CI 1.01–3.97YesYesBienaime et al. (2017) [19]Post10–38 m2nd generation iPTHContinuous845NR54 (42–64)Graft failureContinuous log PTH: HR 3.59 (2.34–5.51)NoYesBleskestad et al. (2014) [20]Post10 w2nd generation iPTHPTH quartiles389NR>60All-cause mortalityPHPT: HR 3.63, 95% CI 0.77–17.12Graft failurePHPT: HR 3.70, 95% CI 1.04–13.10Death-censored graft failurePHPT: HR 4.65, 95% CI 0.54–39.78NoYesCallender et al. (2017) [21]Pre–2nd generation iPTH756 pg/ml or on treatment with VDRA or calcimimetics9131.42%–Graft failureHPT: OR 9.08, 95% CI 2.31–35.71Not reportedNo (different effect measure)Crepeau et al. (2023) [22]Post12 mNR70 pg/ml82460.9%NRAll-cause mortalityPHPT HR 1.35 95% CI 0.96–1.91Graft failurePHPT: HR 1.37, 95% CI 1.04–1.82Normocalcemic HPT: HR 1.50, 95% CI 1.08– 2.08Death-censored graft failurePHPT: HR 1.60, 95% CI 1.02–2.49Normocalcemic PHPT: HR 1.77, 95% CI 1.08–2.90YesYesCrepeau et al. (2024) [23]Pre–NR600 pg/ml113612.5%–All-cause mortalityHPT: HR 0.71, 95% CI 0.47–1.09Death-censored graft failureHPT: HR 1.76, 95% CI 1.16–2.65YesYesDoi et al. (2020) [24]Post>12m2nd generation iPTHContinuous263NR38 (29–52)Graft failure + 50% eGFR declineContinuous log PTH: HR 1.60, 95% CI 1.19–2.14YesYesEgli et al. (2024) [11]PostUp to 24mNR65 pg/ml1212Overall 61.4%, Hypercalcemic 8.5%∼51All-cause mortalityPHPT: HR 2.61, 95% CI 1.62–4.31Hypercalcemic PHPT: HR 2.08, 95% CI 1.05–4.17)Graft failurePHPT: HR 3.88, 95% CI 2.43–6.19Hypercalcemic PHPT: HR 2.22 95% CI 1.19–4.17Death-censored graft failure:PHPT: HR 32.67, 95% CI 4.53–235.51Hypercalcemic PHPT: HR 3.23, 95% CI 0.88–12.5Not reportedYesHanson et al. (2026) [25]Post>6 mNR65 pg/ml + Ca > 2.6 mmol/l887Hypercalcemic 16.4%50.0 (37.2–62.2) vs 51.0 (40.8–65.0)Graft failureHypercalcemic PHPT: HR 2.5, 95% CI 1.78–3.53YesYesIsakov et al. (2020) [9]Post3 mNR150 pg/ml40537.8%53.4 ± 18.3 vs 48.8 ± 21.0All-cause mortalityPHPT: HR 2.1, 95% CI 1.0, 4.3Graft failurePHPT: HR 1.83, 95% CI 1.05, 3.2YesYesLou et al. (2015) [26]PostUp to 24 mNR72 pg/ml160943.1%NRGraft failure: worst survival in Kaplan–Meier for PHPTNoNo (effect measure not provided))Lu et al. (2020) [27]PostNot reported2nd generation iPTHContinuous600NR∼43Graft failureContinuous PTH HR 1.001, 95% CI 1.000–1.003YesYesMolinari et al. (2023) [28]PostUp to 12 m2nd generation iPTH36 pg/ml871Overall 77%,Hypercalcemic 15%64.3 ± 23.6Graft failureContinuous PTH (12 m) OR 2.82, 95% CI 1.23–6.45Hypercalcemic PHPT (12 m) OR 2.27, 95% CI 0.93–5.55YesNo (different effect measure)Molnar et al. (2012) [29]Pre–1st generation PTH800 pg/ml10.107NR–All-cause mortalityHPT: HR 1.1, 95% CI 0.79–1.55Death-censored graft-failureHPT: HR 1.28, 95% CI 0.87–1.88YesYesOkada et al. (2022) [30]Post12 m2nd generation iPTH80 pg/ml892Normocalcemic 44.7%46.2 (39.2–53.6) vs 42.1 (34.7–50.4)Death-censored graft failureNormocalcemic PHPT: HR 1.94, 95% CI 1.27–2.99YesYesOkada et al. (2023) [31]Post12 m2nd generation iPTH80 pg/ml34044.7%73.4 (67.7–83.2) vs 71.0 (63.9–79.4)Death-censored graft failurePHPT: HR 3.05, 95% CI 1.06–8.79YesYesOkada et al.(2025) [32]Post12m2nd generation iPTH80 pg/ml1264Overall 54.6%,Hypercalcemic 8.6%∼44Death-censored graft failureNormocalcemic PHPT: HR 1.37, 95% CI 0.87–2.15Hypercalcemic PHPT: HR 4.25, 95% CI 1.75–10.35YesYesPihlstrøm et al. (2015) [10]Post>5 y2nd generation iPTH65 pg/ml1840NRsCreatinine 145 ± 53 μmol/lAll-cause mortalityPHPT: HR 1.46, 95% CI 1.12–1.92Death-censured graft failurePHPT: HR 1.85, 95% CI1.41–2.42YesYesPrakobsuk et al. (2017) [33]Post>12 m2nd generation iPTH90 pg/ml23764.4%64.6 ± 17.4Graft failureContinuous log PTH HR 1.91, 95% CI 1.19–3.06YesYesRodrigues et al. (2024) [34]Pre–Not reported771 pg/mL15767.6%–All-cause mortalityPHPT HR 0.56, 95% CI 0.31–1.03Death-censored graft failurePHPT: HR 1.01, 95% CI 0.45–2.25YesYesRoodnat et al. (2006) [35]Pre–2nd generation iPTHContinuous397NR–Death-censored graft failureContinuous PTH: HR 1.008, 95% CI 1.002–1.015YesNo (different exposure for pre-transplant analysis)Wang et al. (2023) [36]PostNot reportedNot reporter88 pg/ml or calcimimetic use1554Overall 74.9%,Hypercalcemic 21.1%sCreatinine 1.4 mg/dlGraft failurePHPT: HR 4.0, 95% CI 1.5–10.6Normocalcemic PHPT: HR 3.6, 95% CI 1.5–8.7Hypercalcemic PHPT: HR 4.7, 95% CI 1.5–15.8YesYesWolf et al. (2011) [37]PostNot reported2nd generation iPTHContinuous984NR51 ± 21Composite graft failure + deathContinuous log PTH: HR 1.09, 95% CI 0.96, 1.25YesYesKTx, kidney transplantation; PTH, parathyroid hormone; HPT, hyperparathyroidism; PHPT, persistent post-transplant hyperparathyroidism; iPTH, intact parathyroid hormone; N, sample size; NR, not reported; m, months; w, weeks; y, years; VDRA, vitamin D receptor activator; eGFR, estimated glomerular filtration rate; HR, hazard ratio; OR, odds ratio; CI, confidence interval. Characteristics of the studies included in the systematic review. KTx, kidney transplantation; PTH, parathyroid hormone; HPT, hyperparathyroidism; PHPT, persistent post-transplant hyperparathyroidism; iPTH, intact parathyroid hormone; N, sample size; NR, not reported; m, months; w, weeks; y, years; VDRA, vitamin D receptor activator; eGFR, estimated glomerular filtration rate; HR, hazard ratio; OR, odds ratio; CI, confidence interval. The overall results and the pooled HR with 95% CI (both the conventional approach and the Hartung–Knapp sensitivity analysis) are presented in Supplementary Fig. S2. Pre-transplant HPT In an analysis restricted to studies reporting pre-transplant HPT in relation to all-cause mortality and DCGF, three studies were included. Pre-transplant HPT showed no pooled association with all-cause mortality (HR 0.80, 95% CI 0.54–1.18; P = 0.264), with moderate heterogeneity (τ² = 0.069; I² = 58.4%; Q = 4.81, df = 2, P = 0.090). Results were also examined using Hartung–Knapp confidence intervals (HR 0.80, 95% CI 0.34–1.86; P = 0.375) (Fig. 2A). Forest plot of the association between categorically defined pre-transplant HPT and all-cause mortality (A) and death-censored graft failure (B). By contrast, using the classical CI approach, pre-transplant HPT was associated with an increased pooled risk of DCGF (HR 1.42, 95% CI 1.07–1.87; P = 0.013), with low heterogeneity (τ² = 0.004; I² = 0.2%; Q = 2.00, df = 2, P = 0.367) (Fig. 2B). However, in sensitivity analyses applying Hartung–Knapp, confidence intervals widened and the pooled association was no longer statistically significant (HR 1.42, 95% CI 0.79–2.56; P = 0.124) (Supplementary Fig. S2B). Post-transplant Persistent HPT In analyses restricted to studies using a categorical definition of overall Persistent HPT and reporting all-cause mortality (Fig. 3A), five studies were included. Persistent HPT was associated with higher mortality (HR 1.74, 95% CI 1.29–2.35; P = 0.0003). Between-study heterogeneity was moderate (τ² = 0.050; I² = 41.4%), while the Cochran Q test was not statistically significant (Q = 6.82, df = 4, P = 0.146). Using Hartung–Knapp confidence intervals, the association remained statistically significant (HR 1.74, 95% CI 1.16–2.62; P = 0.0197). Forest plot of the association between categorically defined Persistent post-transplant HPT with all-cause mortality (A), overall graft failure (B), and death-censored graft failure (C). For overall graft failure, we pooled data from nine studies. Persistent HPT was associated with an increased risk of graft failure (random-effects HR 2.16, 95% CI 1.65–2.83; P < 0.0001), with moderate-to-substantial heterogeneity (τ² = 0.085, I² = 56.4%) and a significant Q test (Q = 18.33, df = 8, P = 0.018) (Fig. 3B). Results were consistent under Hartung–Knapp (HR 2.16, 95% CI 1.60–2.90; P = 0.0003). When restricting the analysis to studies reporting only DCGF (Fig. 3C), seven studies were included. Persistent HPT remained significantly associated with DCGF (HR 1.92, 95% CI 1.59–2.33; P < 0.0001), with negligible between-study variance (τ² < 0.0001) and moderate I² (40.1%). The Q test was not statistically significant (Q = 10.02, df = 6, P = 0.124), and the association persisted with Hartung–Knapp intervals (HR 1.92, 95% CI 1.42–2.61; P = 0.0020). Sensitivity and exploratory analyses In sensitivity analyses by Persistent HPT phenotype (Fig. 4), normocalcemic Persistent HPT was associated with overall graft failure across four studies (HR 1.66, 95% CI 1.33–2.06; P < 0.0001), with minimal between-study variance (τ² < 0.0001) and low-to-moderate heterogeneity (I² = 34.0%; Q = 4.55, df = 3, P = 0.208); results remained significant under Hartung–Knapp (HR 1.66, 95% CI 1.07–2.57; P = 0.0347). Forest plot of the association between categorically defined persistent post-transplant normocalcemic HPT (A) and hypercalcemic HPT (B) with graft failure. For hypercalcemic HPT, four studies showed a stronger association with overall graft failure (HR 2.67, 95% CI 2.02–3.52; P < 0.0001), with negligible between-study variance and very low heterogeneity (τ² <0.0001; I² = 0.0%; Q = 2.40, df = 3, P = 0.393). Association was consistent with Hartung–Knapp intervals (HR 2.67, 95% CI 1.79–3.99; P = 0.004). Finally, in an exploratory analysis of studies treating PTH as a continuous exposure with overall graft failure (Fig. 5), the classic random-effects model suggested an association (HR 1.59, 95% CI 1.02–2.47; P = 0.048) but with very high heterogeneity (τ² = 0.226; I² = 92.2%; P < 0.0001); under Hartung–Knapp, the estimate was unchanged but no longer statistically significant (HR 1.59, 95% CI 0.85–2.98; P = 0.108). Forest plot of the association between continuous log-transformed PTH and overall graft failure. DISCUSSION To our knowledge, this is the first systematic review and meta-analysis that analyzed the association between pre- and post-transplant HPT and graft and patient outcomes in KTRs. Pre-transplant HPT was not associated with post-transplant mortality. For DCGF, the point estimate suggested a higher risk, but this result was based on only three studies and became imprecise after Hartung–Knapp adjustment; therefore, the evidence should be considered suggestive rather than definitive. Interpretation is further limited by the fact that severe pre-transplant HPT was infrequent and variably defined across studies, underscoring the lack of a standardized threshold for clinically relevant HPT severity in patients with end-stage kidney disease (ESKD). Conversely, post-transplant Persistent HPT was robustly and consistently associated with worse hard clinical outcomes after kidney transplantation, with statistically significant effects in the primary analyses and findings that were confirmed across sensitivity analyses, strengthening the inference that Persistent HPT is a clinically relevant risk marker in this setting. When HPT was defined categorically, Persistent HPT was associated with higher risks of all-cause mortality (HR 1.74) and graft failure (HR 2.16), with broadly consistent Hartung–Knapp results. The association also extended to DCGF (HR 1.92), supporting a link with graft outcomes beyond the competing risk of death. Phenotype-specific analyses suggested a gradient of risk across calcium-defined forms: normocalcemic Persistent HPT was associated with a moderate increase in graft failure risk (HR 1.66), whereas hypercalcemic HPT showed a larger effect (HR 2.67). Notably, phenotype stratification substantially reduced between-study variance and heterogeneity, suggesting that calcium-based phenotyping captures clinically meaningful differences and may explain part of the variability seen with overall categorical definitions. Several studies have linked Persistent HPT to higher mortality. In 1840 KTRs, Pihlstrøm et al. found an association with all-cause mortality, but not with major cardiovascular events [10]. More recently, Hanson et al. reported that, in a cohort of 887 KTRs, patients with hypercalcemic Persistent HPT had a 20-year post-transplant survival of 35%, compared with 57% in those without Persistent HPT [25]. The mechanisms underlying this excess mortality remain unclear. However, in CKD, HPT has been associated with increased mortality through pathways involving elevated Fibroblast Growth Factor 23 (FGF-23) and reduced soluble Klotho and vitamin D levels [38–40], which may promote vascular calcification, left ventricular hypertrophy, and arrhythmias [41–43]. In the general population, Primary HPT has also been associated with higher all-cause and cardiovascular mortality, highlighting the systemic burden of sustained PTH excess [44]. This may be particularly relevant in the transplant setting, where hypercalcemic Persistent HPT shares important biochemical and clinical features with Primary HPT. Furthermore, this meta-analysis confirmed a strong association between Persistent HPT and graft failure, both overall and death-censored, indicating that the signal is not explained solely by competing mortality and is consistent with mechanisms of allograft injury. Several studies have shown that elevated PTH levels, with or without hypercalcemia, are associated with a higher risk of graft failure. In a cohort of 824 KTRs, Crepeau et al. reported a 1.37-fold higher risk of all-cause graft loss in patients with Persistent HPT; when stratified by PTH level, the risk was 2.46-fold higher in those with PTH > 300 pg/ml than in those with PTH < 70 pg/ml [20]. Several mechanisms may underline this association. Gwinner et al. and Evenepoel et al. reported an increased risk of nephrocalcinosis in patients with Persistent HPT, characterized by calcium crystal deposition in the tubulointerstitium [13, 14]. Gwinner et al. also observed poorer graft outcomes at 1 year in patients with calcium crystal deposits and elevated PTH, whereas Evenepoel et al. did not confirm this association [13, 14]. Other studies suggest that PTH may promote graft fibrosis, possibly through inflammation related to nephrocalcinosis [15]. Experimental data further support a direct profibrotic effect, as PTH can induce epithelial-to-mesenchymal transition in human proximal tubular cells via the Wnt/β-catenin pathway [43, 45]. A vascular mechanism has also been proposed, whereby PTH may affect glomerular hemodynamics through vascular receptors in renal arterioles, stimulate renin release, and increase vascular stiffness [17, 46, 47]. In phenotype-stratified sensitivity analyses, both normocalcemic and hypercalcemic HPT were associated with a higher risk of graft failure. Normocalcemic HPT increased risk by 66% with low heterogeneity and remained significant under Hartung–Knapp adjustment, supporting the robustness of the estimate. Hypercalcemic HPT showed a larger effect, with an approximately three-fold higher risk, negligible heterogeneity, and similarly stable Hartung–Knapp results. Overall, both calcium-defined phenotypes appear to convey clinically meaningful prognostic information for long-term graft outcomes. These data renew the discussion on the role of hypercalcemia in graft dysfunction among patients with Persistent HPT. In a retrospective cohort of 1212 KTRs, Egli et al. found that hypercalcemic HPT was associated with poorer graft outcomes, whereas normalization of both calcium and PTH was linked to significantly better DCGF and overall patient survival [11]. By contrast, Crepeau et al. did not identify a specific role for hypercalcemia in KTRs with Persistent HPT [22]. In this context, assessment should extend beyond hypercalcemia alone and also consider hypercalciuria, phosphate balance, and the calcium-sensing receptor (CaSR), which may help better define the HPT phenotype. Persistent HPT may present either as a hypercalciuric phenotype or as a relatively hypocalciuric, calcium-raising phenotype; the latter can resemble familial hypocalciuric hypercalcemia and may reflect altered calcium sensing and/or increased peripheral responsiveness to PTH (Fig. 6) [48, 49]. Since CaSR is expressed both in the parathyroid glands and along the nephron, where it regulates PTH secretion and renal calcium handling, the phenotype of Persistent HPT may depend not only on parathyroid autonomy, but also on dysregulated CaSR signaling, potentially reflecting a shifted calcium-sensing set point [50]. Proposed pathophysiological framework linking PTH excess to calcium phenotypes and potential allograft injury. Persistent HPT increases bone turnover, tubular Ca²⁺ handling, and intestinal Ca²⁺ absorption (via 1,25(OH)₂D), promoting hypercalcemia and interacting with RAAS activation (renin/Ang II) and intrarenal hemodynamics (↑ AA and EA resistance). Differences in CaSR signaling may delineate a hypercalciuric phenotype (nephrocalcinosis, tubulo-interstitial fibrosis) vs a hypocalciuric phenotype (vascular calcification, vasoconstriction). Abbreviations: PTH, parathyroid hormone; HPT, hyperparathyroidism; Ca²⁺, calcium; 1,25(OH)₂D, 1,25-dihydroxyvitamin D; RAAS, renin–angiotensin–aldosterone system; Ang II, angiotensin II; AA, afferent arteriole; EA, efferent arteriole; CaSR, calcium-sensing receptor. Created in https://BioRender.com. CaSR has been shown to play a pivotal role in controlling extracellular Ca²⁺ levels in a PTH-independent manner, through regulation of Ca²⁺ excretion in the thick ascending limb of the loop of Henle [50]. Clinically, these phenotypes may imply different injury pathways: a hypercalciuric pattern may increase intratubular calcium load and predispose to nephrolithiasis/nephrocalcinosis and tubulointerstitial injury, whereas a relatively hypocalciuric, calcium-raising pattern may be more closely related to systemic and hemodynamic consequences of hypercalcemia (e.g. renal vasoconstriction). Interpretation remains constrained because most studies have not concurrently assessed urinary calcium excretion and bone turnover, thereby limiting discrimination and risk stratification. Beyond their prognostic relevance, these findings raise the question of whether Persistent HPT is a modifiable post-transplant risk factor. Interventional evidence in KTRs remains limited: randomized studies have mainly addressed biochemical and skeletal endpoints, while effects on graft and patient survival remain unproven [51, 52]. Consequently, evidence on hard outcomes comes largely from retrospective cohorts and is inconsistent. Okada et al. reported that untreated hypercalcemic HPT was independently associated with a higher risk of graft failure, whereas recipients treated with parathyroidectomy (PTX) or calcimimetics within the first post-transplant year had graft survival comparable to patients without HPT [32]. Wang et al. similarly found lower risks of all-cause graft failure and graft failure after medical or surgical correction of hypercalcemic HPT [36]. By contrast, Hanson et al. found no clear long-term benefit of PTX or calcimimetics [25]. Overall, observational data suggest a potentially modifiable risk factor, but the evidence remains inconclusive. Adequately powered, ideally randomized, trials are needed to determine whether active correction of Persistent HPT improves long-term graft and patient outcomes and to identify the phenotypes most likely to benefit. Proposed clinical roadmap Given the high prevalence of post-transplant HPT recurrence, implementing a clinical flowchart for transplant candidates may help mitigate this risk (Fig. 7). The KDIGO guidelines on evaluating and managing candidates for kidney transplantation recommend delaying transplantation in patients with severe HPT until adequate treatment is achieved, in accordance with CKD-MBD guidelines [53, 54]. However, there is no clear consensus on what defines adequate control of SHPT in waitlisted patients. The 2017 KDIGO guidelines suggest maintaining PTH levels between two and nine times the upper limit of normal, but this range is rather nonspecific, as it does not consider calcium levels, therapeutic response, or assay variability [54]. Moreover, a standardized definition of Persistent HPT is still lacking. Recent controversies emphasize the need for future guidelines to distinguish between Persistent and recurrent secondary HPT after transplantation, given their distinct biochemical profiles and pathophysiological mechanisms [2]. Proposed clinical flowchart for the management of hyperparathyroidism in kidney transplant candidates and recipients. The left panel outlines a pragmatic approach for kidney transplant candidates based on the evaluation of CKD–MBD-related risk factors and therapies. Persistently elevated PTH levels (>300 pg/ml) should be interpreted in the context of calcium and phosphate levels, as well as the presence of secondary HPT-related symptoms or complications (bone pain, pruritus, and EPO-resistant anemia). Patients receiving high doses of vitamin D receptor activators/vitamin D analogues should be carefully assessed, as these treatments may influence biochemical parameters. Imaging (neck ultrasound ± CEUS and scintigraphy, followed, when indicated, by CT or choline PET) supports the identification of enlarged or ectopic parathyroid glands, and management should be discussed within a multidisciplinary team. The right panel summarizes the management of kidney transplant recipients with persistent hyperparathyroidism, defined as PTH levels >1.5–2× the upper limit of normal, after the exclusion of other metabolic abnormalities (e.g. vitamin D deficiency). Hypercalcemia is stratified as mild (10.5–11.9 mg/dl), moderate (12.0–13.9 mg/dl), or severe (≥14 mg/dl, or hypercalcemia not responsive to medical treatment). Severe hypercalcemia may prompt consideration of parathyroidectomy, whereas mild-to-moderate cases may be managed with calcimimetics and close follow-up, with reassessment after 6–12 months. Abbreviations: CKD–MBD, chronic kidney disease–mineral and bone disorder; PTH, parathyroid hormone; HPT, hyperparathyroidism; SHPT, secondary hyperparathyroidism; EPO, erythropoietin; VDRA, vitamin D receptor activator; CEUS, contrast-enhanced ultrasound; CT, computed tomography; PET, positron emission tomography; ULN, upper limit of normal. Created in https://BioRender.com. Given the current uncertainties and the lack of standardized criteria, we propose a pragmatic clinical flowchart to support decision-making in this setting. The first step should be a comprehensive patient assessment, including biochemical parameters (PTH, calcium, phosphate, vitamin D), temporal trends, clinical manifestations, and ongoing treatments. Particular attention should be paid to risk factors for Persistent HPT, such as markedly elevated pre-transplant PTH levels, prolonged dialysis vintage, and conditions associated with longer waiting times, including retransplant candidates and highly sensitized patients likely to remain on the waiting list for several years [22, 25]. PTH levels are central to risk stratification; in particular, patients with values above the KDOQI target (>300 pg/ml, notably lower than the KDIGO range) should be carefully evaluated, together with severe hypercalcemia or hyperphosphatemia, calcimimetic use, and possible rebound in PTH or calcium after calcimimetic withdrawal. Once high-risk patients have been identified, imaging should be performed to confirm parathyroid enlargement. First-line evaluation may include scintigraphy and neck ultrasound, with contrast-enhanced ultrasound as an adjunct. Parathyroid gland size is also relevant, as a diameter >8–10 mm has been associated with more severe HPT and treatment resistance [55–57]. Second-line imaging, such as neck CT or choline PET, may be considered particularly in patients with long dialysis vintage, in whom ectopic glands are more common. Finally, patients should be evaluated by a multidisciplinary team comprising a nephrologist (with bone and transplant expertise), an endocrinologist, a radiologist, and a surgeon for optimal management of pre-transplant HPT and to reduce the risk of recurrence. Patients with risk factors, enlarged parathyroid glands on imaging, and poor control of HPT should be evaluated for PTX [7]. When HPT recurs after transplantation, with PTH levels >1.5–2× the upper limit of normal after exclusion of other metabolic abnormalities such as vitamin D deficiency, and with or without hypercalcemia or hypercalciuria, it is important to consider evidence from the literature on treatment with calcimimetics or PTX in the post-transplant setting. Although patients treated with PTX experienced a reduction in graft function after surgery, this was not confirmed after 1 year [51]. Additionally, PTX showed better results in controlling HPT and increasing bone mineral density. In these cases, it could be useful to consider surgery in patients with recurrence of HPT associated with severe hypercalcemia, and in other cases, apply a strict follow-up to evaluate the evolution of Persistent HPT and reconsider PTX after 6–12 months post-transplant [58]. Limitations and strengths This meta-analysis has several limitations. First, exposure definitions were heterogeneous across studies, with HPT defined using different PTH thresholds, variable incorporation of calcium status, and, in some cohorts, treatment-based criteria. Second, the timing of PTH assessment varied widely, and definitions requiring persistence beyond 12–24 months may be prone to landmark/selection (“immortal time”) bias, as patients must survive with a functioning graft long enough to be classified as having Persistent HPT. Third, the included studies were almost exclusively retrospective observational cohorts, often single-center or registry-based, with substantial differences in covariate adjustment, increasing susceptibility to selection bias, residual confounding, and exposure misclassification. Because PTH is closely linked to kidney function, residual confounding and reverse causation remain major concerns, particularly for graft outcomes, as impaired graft function may itself contribute to Persistent PTH elevation. Interpretation is further complicated by the fact that most studies did not clearly distinguish Persistent HPT from de novo secondary HPT related to incomplete recovery or later graft dysfunction. Finally, analyses of continuous PTH were limited by heterogeneous effect scaling and possible assay-related variability, likely contributing to substantial heterogeneity and less robust estimates under Hartung–Knapp intervals. This study also has several strengths. We used a clinically oriented and methodologically explicit framework, separating analyses by PTH operationalization (categorical vs continuous), endpoint definition, and HPT phenotype, thereby improving interpretability and reducing the risk of pooling non-comparable effect measures. We applied contemporary random-effects methods (REML) and systematically assessed robustness using Hartung–Knapp confidence intervals, providing more conservative inference in the setting of few studies. Importantly, the main associations for categorical post-transplant HPT were consistent across mortality and graft outcomes, including analyses restricted to death-censored graft failure, supporting internal coherence. Although heterogeneity in PTH assessment and HPT definitions was a limitation, it may also support external validity, as the association between Persistent HPT and adverse outcomes remained broadly consistent across diverse thresholds, assays, and non-standardized terminology, reflecting real-world nephrology practice, where assay harmonization and guideline-endorsed nomenclature are still lacking. CONCLUSIONS In this systematic review and meta-analysis, pre-transplant HPT showed no clear association with post-transplant mortality while the evidence for graft outcomes remained limited and should be interpreted cautiously. By contrast, Persistent HPT was consistently associated with worse hard outcomes after kidney transplantation, including graft failure and mortality, across heterogeneous definitions and clinical settings. Although pre-transplant HPT does not appear to show a clear association with post-transplant clinical outcomes, it is undeniable that Persistent HPT shows a clinical and pathophysiological “continuum” with pre-transplant HPT and therefore requires an appropriate pharmacological and surgical approach in the renal transplant candidates [7, 53]. These findings support closer surveillance and a proactive management of mineral–bone abnormalities after transplantation. From a research perspective, the priority is to move beyond risk association and develop a standardized framework for Persistent HPT after kidney transplantation. Future guidelines should establish harmonized definitions incorporating PTH levels, assay characteristics, timing of assessment and mineral metabolism parameters (including calcium, calciuria, and phosphaturia). Prospective studies should address key unanswered questions, including optimal thresholds, timing of intervention and phenotype-specific risk (e.g. normocalcemic vs hypercalcemic and hypercalciuric forms), as well as the underlying mechanisms linking HPT to graft dysfunction. Given current limitations, future investigations should prioritize prospective cohorts, pragmatic trials and randomized studies with time-varying exposure modelling and phenotype-based stratification. Clinically relevant endpoints should extend beyond biochemical control to include overall graft failure, DCGF, nephrocalcinosis, fractures, bone mineral density, cardiovascular events, overall mortality and patient-reported outcomes, to determine whether Persistent HPT is a modifiable therapeutic target rather than solely a prognostic marker. Supplementary Material Contributor Information Daniele Vetrano, Department of Medical and Surgical Science (DIMEC), Alma Mater Studiorum—University of Bologna, Bologna, Italy. Simona Barbuto, Nephrology, Dialysis and Kidney Transplant Unit, IRCCS Azienda Ospedaliero-Universitaria di Bologna, Bologna, Italy. Francesco Aguanno, Department of Medical and Surgical Science (DIMEC), Alma Mater Studiorum—University of Bologna, Bologna, Italy. Paolo Mastromauro, Department of Medical and Surgical Science (DIMEC), Alma Mater Studiorum—University of Bologna, Bologna, Italy. Valeria Grandinetti, Nephrology, Dialysis and Kidney Transplant Unit, IRCCS Azienda Ospedaliero-Universitaria di Bologna, Bologna, Italy. Giorgia Comai, Department of Medical and Surgical Science (DIMEC), Alma Mater Studiorum—University of Bologna, Bologna, Italy; Nephrology, Dialysis and Kidney Transplant Unit, IRCCS Azienda Ospedaliero-Universitaria di Bologna, Bologna, Italy. Gaetano La Manna, Department of Medical and Surgical Science (DIMEC), Alma Mater Studiorum—University of Bologna, Bologna, Italy; Nephrology, Dialysis and Kidney Transplant Unit, IRCCS Azienda Ospedaliero-Universitaria di Bologna, Bologna, Italy. Giuseppe Cianciolo, Nephrology, Dialysis and Kidney Transplant Unit, IRCCS Azienda Ospedaliero-Universitaria di Bologna, Bologna, Italy. CONFLICT OF INTEREST STATEMENT None declared. DATA AVAILABILITY STATEMENT No new data were generated or analyzed in support of this research. References Associated Data Supplementary Materials Data Availability Statement No new data were generated or analyzed in support of this research.